Human Factors Assessment of Pedestrian Roadway Crossing Behavior

Data collection validation

Video recording pedestrian crossings is not a common method for assessing pedestrian behaviors. Further, to the authors’ knowledge, no other studies have attempted to use existing traffic cameras/CCTV to assess pedestrian behavior. As such, it was deemed important to assess the validity of camera footage as a data collection methodology.

Researchers manually scored pedestrian crossing behaviors at Location 4 (13th Street and I Street Northwest) and Location 5 (14th Street and H Street Northwest) onsite. These two locations were selected for their physical proximity to one another. Researchers recorded the pedestrians’ crossings and their interactions with vehicles (described in more detail in the next section) over three 15-min periods at both locations. The in-vivo recordings were made the morning of February 20, 2012 (President’s Day). Unfortunately the camera recording Location 5 was facing the incorrect direction on this day. As a result, there is no video data to compare with the live, onsite scoring. However, the onsite pedestrian crossing scoring was compared with the DDOT video at Location 4.

At Location 4, there was a 100-percent agreement in the classification of the pedestrian crossings for both the first and last 15-min segments (18 and 21 crossings, respectively). However, in the second 15-min session, there was a single discrepancy; the onsite coding resulted in five total pedestrian crossings, and the video coding resulted in four crossings. The categorization of the four video crossings was the same as those in the in-vivo coding. This 98-percent agreement between the two coding methodologies provides evidence that video coding is a reliable methodology for coding and characterizing pedestrian crossing behaviors.

Video data coding

Many different types of pedestrian crossings and pedestrian interactions with vehicles in, and along, the roadway can be recorded and classified. Although it is difficult to code pedestrian crossings in an exhaustive manner, the current study sought to record enough information to interpret general crossing behaviors. For each pedestrian crossing, multiple factors were recorded. The following subsections describe each of these factors and how data were coded.

Crossing Factors

Data were coded between the marked intersection closest to the DDOT camera and a far marked intersection (as denoted in each of the location descriptions). Pedestrian crossings were not counted in the far marked intersections. Only completed crossings were recorded and included in data analysis. In other words, if a person walked partially into the roadway to hail a cab, it was not included as a pedestrian crossing.

Location/Crossing Area

The location where pedestrians crossed the road was classified in one of three categories:

·         Marked Intersection. This included any crossing that occurred in any portion of the marked crosswalk at the intersection closest to the DDOT camera. Crossings within one car length of the marked crosswalk were also included in this category. This is simply because some of the marked intersections contained enough pedestrian traffic that all persons could not comfortably cross within the marked crosswalk.

·         Unmarked Intersection. This was any crossing at an intersection without a marked crosswalk. This type of crossing only applied to Location 3.

·         Unmarked Non-Intersection. This type of crossing refers to any pedestrian crossing that occurred at a non-intersection location without a marked crossing.

Traffic

The status of vehicular traffic was also noted.

·         With Traffic. This is a crossing that does not conflict with traffic flow. An example is a crossing in the crosswalk with the walk sign activated.

·         Against Traffic. This is a crossing that conflicts with traffic flow. An example is crossing in the crosswalk while oncoming traffic has a green light and the don’t walk sign is activated. (The don’t walk sign must be in the steady state to be counted as in the don’t walk phase.)

·         Traffic Flow Change. This is a crossing that starts either with or against traffic, but ends with the opposite traffic flow. An example of this type of crossing is one in which a pedestrian starts crossing the street with the don’t walk sign activated and completes the crossing with the walk sign activated.

Yielding Behavior

Yielding behavior of both pedestrians and vehicles was recorded. A note on where the yielding occurred and the light status was recorded. Vehicle or pedestrian right of way can be established using this information. (Note that right-of-way laws vary by State. However, in the present context, the legality of the crossing is not considered.)

·         Pedestrian Yielding. This was recorded when a pedestrian yielded to a vehicle. An example is a pedestrian pausing in the roadway to allow a turning vehicle to complete a turn, prior to the pedestrian completing a crossing.

·         Vehicle Yielding. This was recorded when a vehicle yielded to a pedestrian. An example is a vehicle pausing mid-turn to allow a pedestrian to complete the crossing.

Evasive Pedestrian Actions

Actions made by pedestrians to avoid conflict or potential collisions were also recorded. A distinction between evasive actions in the first half and second half of the crossing was made.

·         Running/Accelerated Walking. This was recorded when a pedestrian either ran or noticeably increased walking speed to avoid collision or potential conflict with a vehicle. Note that pedestrians who ran across the intersection, but were not taking evasive actions (e.g., joggers), were not coded as taking this type of evasive action.

·         Abrupt Stopping. This was recorded when a pedestrian stopped abruptly in the roadway to avoid collision or potential conflict with a vehicle.

·         Directional Change. This was recorded when a pedestrian changed travel direction, once in the roadway, to avoid collision or potential conflict with a vehicle. An example of this is a pedestrian stepping into the roadway and then subsequently returning to the curb (direction change) after noting an oncoming vehicle.

Evasive Vehicle Actions

Evasive actions taken by vehicles were also recorded. A distinction between evasive actions taken by the vehicle closest to the pedestrian and a following vehicle (if any) was made.

·         Abrupt Braking (First Vehicle). This was recorded when a vehicle was forced to abruptly brake (not a gradual deceleration) to avoid collision with a pedestrian.

·         Abrupt Braking (Second Vehicle). This was recorded when the action of the vehicle closest to the pedestrian required the next vehicle to brake abruptly to avoid collision.

·         Directional Change (First Vehicle). This was recorded when a vehicle swerved (or otherwise modified direction) to avoid collision with a pedestrian.

·         Directional Change (Second Vehicle). This was recorded when the action of the vehicle closest to the pedestrian required the next vehicle to modify direction to avoid collision.

Other Relevant Information

Other relevant information related to the pedestrian crossings was also recorded. For example, if an ambulance or presidential motorcade interrupted regular traffic flow, this was marked to accommodate potential irregular data coding.